As already said to you, your idea wont work due to the way no other task is performed untill the currently running script completes, ie: infinite loop completes (an unlikely occurance!)
While it might help in shutting down events in some instances of high levels of events triggering causing you to lag out such as a netsplit/join or an attack on a scripted event, I would say it would be far more clever to script protection into those events.
If you want to controll possable flooding problems i would look to a debug session, which deactivates events, however thats quite a big task to craft it correctly.
It nice for those of us running nearly one meg of scripts.
Sounds like your relying on a bulk of other peoples scripts to be getting that higher size, I have found that often scripts compete with each other causing more trouble.
Would also like to see a command that will stop all events/aliases currently running, sort of a control-break via scripts.
Stopall { !debug off | !remote off | timer* off }
^ that of course must get a chance to run, which means the system has at least freed up for the moment that it is run!